The well-known ChatGPT will soon be available on Android, but research indicates that the generative AI has begun to slow down.
It has been reported that ChatGPT will soon be available as an app for Android users on the Play Store. This announcement comes almost after 8 weeks of releasing OpenAI on iOS. Users of Android can “pre-register” on the Google Play Store to receive notifications when the app is ready.
Announcing ChatGPT for Android! The app will be rolling out to users next week, and you can pre-order in the Google Play Store starting today: https://t.co/NfBDYZR5GI
— OpenAI (@OpenAI) July 21, 2023
However, OpenAI has not provided a release date for the app. According to reports, the app will first be available for Android users in the United States, followed by other nations.
What has changed in the app?
- While mobile users may still utilize the web to access ChatGPT, an app will deliver a superior user interface experience.
- Except for voice assistance changes, the Android version app is planned to appear similar to the iOS app.
- The app will also likely include the majority of the functionality seen on the web-based interface, such as the ability to sync chats between devices.
- If the success of this on iOS is any indication, the app will have at least 500,000 downloads in its first week.
- In the first week, iOS users installed the app around 500,000 times.
Also, check out: WormGPT AI is the evil twin of ChatGPT and can generate malware
However, there are a few worries regarding ChatGPT:
GPT-4 is getting worse over time, not better.
— Santiago (@svpino) July 19, 2023
Many people have reported noticing a significant degradation in the quality of the model responses, but so far, it was all anecdotal.
But now we know.
At least one study shows how the June version of GPT-4 is objectively worse than… pic.twitter.com/whhELYY6M4
- According to one research, the most recent version of ChatGPT is getting little bit dumb, sluggish and less reliable than prior versions.
- From March to June 2023, researchers at Stanford and UC Berkeley evaluated multiple versions of ChatGPT on four major tasks: solving math problems, responding to sensitive questions, visual reasoning, and creating code.
- They discovered that ChatGPT-4 could recognize prime numbers with 97.6% effectiveness in the month of March but by June, it could only answer 12 inquiries correctly with an accuracy rate of 2.4%.
- Twitter users speculated that ChatGPT’s low performance was caused by OpenAI’s cost-cutting measures. It’s possible that the corporation isn’t releasing their finest Language models to everyone.
- On the contrary, there are studies that demonstrate ChatGPT can equal the intellect of the top 1%.
- This discovery was found by University of Montana researchers who evaluated ChatGPT in the Torrance Tests of Creative Thinking, which are commonly used to assess creativity.
Can the latest AI match human creativity? New research from #UM researcher Erik Guzik and @UMontanaBiz suggests it's closer than we think.
— University of Montana (@umontana) July 6, 2023
👉 https://t.co/OG7SLPddUG pic.twitter.com/RDCGm0uJfm